No you can't use logic based on old information...
If information changes between variables a language model can't understand that, because it doesn't understand.
If your information relies on x being true, when x isn't true the ai will still say its fine because it doesn't understand the context
Just like it doesn't understand things like not to do something.
I was thinking about this recently... and in the early 2000s for a short time there was this weird chat bot crazy on the internet... everyone was adding them to web pages like MySpace and free hosting sites...
I feel like this has been the resurrection of that but on a whole other level... I don't think it will last it will find its uses but shoving glorified auto suggest down people's throats is not going to end up anywhere helpful...
A LLM has its place in an ai system... but without having reason its not really intelligent. Its like how you would decide what to say next in a sentence but without the logic behind it
He's going to invade Iceland because thats what this is really about... hes just so dumb to know the difference
People who ask these questions generally are the type of people who treat others badly because they believe they are better than them... so they expect to be treated how they treat people
I've done the math on this before in another posting...and you basically need like 30 balloons for an average dog poop for this to work...
"And by the way, I'm a fan of Denmark, too, I have to tell you. ... But, you know, the fact that they had a boat land there 500 years ago doesn't mean that they own the land. I'm sure we had lots of boats go there also," he said.
It's not smart... People need to understand this... its not making choices based on logic or data or facts.
It has been given data that says trump pedophile and then the image processing part finds the thing labeled trump and removes him based on the label... thats it
Can he do something productive like fixing Arkansas and Kansas to actually make sense....
/s
Thing about it currently is, these "ai" tools built into browsers are a huge security risk... they have access to basically everything the browser does. A prompt on some web page can tell the ai to go do what ever the he'll the attacker wants? Your bank login Details, sure why not! Your saved passwords built into Firefox sure why not... like no thank you
Yeah, every time trump says its fake news... and then you know does nothing to try to improve reporting to be more factual... I remember how much of a shit head he is...
Yeah and they just did it again...